(a). Field of the Invention
The present invention relates to a method and a device for processing digital image, more particularly, to a method and a device that will separate the image's foreground data from its background data before coding the image into a coded image file.
(b). Description of the Prior Arts
Conventionally, because the amount of data of a digital image file is huge, large bandwidth and high performance are needed for a processing device when handling these image data without compression. The common technologies used for compressing the digital image data include GIF, JPG. PIC and so forth. Some of the technologies may reduce the resolution of the image data, others don't. Usually, the compression ratio of the image data with reduced resolution is much higher than that without any reduction. Nevertheless, no matter if the resolution is reduced or not, all of these conventional compressing technologies always compress the whole image data in one shot. No pre-handling process is performed on the image data before starting the compressing process. Therefore, the compression ratio will be limited in order to obtain a compressed image file without reducing the resolution. However, in real practice, the foreground data of a regular image may only be in a certain portion of the whole image, for example, may only be 20% of it, and the rest is the background data. Therefore, if the foreground data and the background data can be separated in advance before performing the compressing process of the image data, the size of the compressed image file will be extremely and efficiently reduced.
In addition, due to the limitation of sensitivity and the characteristic of Analog-to-Digital Converter (ADC), sometimes an image sensor may still output an uneven result of scan even when the image being scanned is merely a background image with only one color. The quality of the scanned image is seriously impacted.